This disclosure relates to a device for detecting at least one analyte in a bodily fluid by means of at least one test element and preferably by means of at least one lancet with a capillary. This disclosure furthermore relates to a method for recognizing an evaluation region of a device for detecting at least one analyte in a bodily fluid. Such devices and methods are used, in particular, in diagnostics for qualitatively or quantitatively detecting one or more analytes, for example one or more metabolites such as e.g. blood glucose, in bodily fluids such as e.g. blood or interstitial fluid.
The prior art has disclosed a multiplicity of devices for detecting at least one analyte in a bodily fluid. Here, use is generally made of test elements which have at least one test chemical. This test chemical contains at least one detection reagent which, when it comes into contact with the at least one analyte, carries out an analyte-specific reaction, which can for example be detected by electrochemical and/or optical means.
In addition to individual systems, in which the sample of bodily fluid is obtained and analyzed separately, integrated systems in particular have prevailed in recent times. By way of example, integrated systems for determining the blood glucose are composed of means for obtaining blood and means for determining glucose. In the step of obtaining blood, some systems make use of a flat lancet with a semi-open microcapillary, in which the capillary typically has a width of 120 μm and a length of 4 mm. After obtaining blood by means of a piercing process, for example into a finger tip, an ear lobe or a forearm, the blood taken up into the capillary is often transferred onto a test field of the test element by virtue of causing the lancet to approach said test field, for example by being pressed onto the latter. As a result, there likewise is the creation of an approximately 120 μm wide print of the capillary on the test element, for example a strip-shaped test element, which changes depending on the blood glucose content and depending on the test chemical recipe. In the case of optical systems, this change generally consists of a local change in color, which can be measured by reflectance photometry. In principle, details of this method are sufficiently well known from the literature.
Here, in principle, there is a problem in that measuring the discoloring of the test fields by means of non-spatially-resolved sensors, i.e. by means of for example a single photodiode, is problematic in that the position of the grayscale discoloring or discoloring on the test field in this case would have to be captured very precisely by mechanical means and with low mechanical and/or optical tolerances. This is difficult to achieve in the case of a lancet that can move for piercing and for obtaining blood. By way of example, if a tolerance in the lateral position of less than 10% is to be achieved, the positional tolerance in the case of a capillary with a width of 120 μm must not exceed approximately ±10 μm; this is a significant challenge from a mechanical point of view.
It is for this reason that the use of a spatially resolved detector, e.g. a CMOS camera, has been proposed a number of times. By way of example, U.S. Pat. No. 6,847,451 B1 describes devices and methods for determining the concentration of an analyte in a physiological sample. Here, use is made of at least one light source and at least one detector array, as well as of means for determining whether a sufficient amount of sample is present on a plurality of different surfaces. Inter alia, it is proposed here to use a CCD array as detector array. Alternatives to the spatially resolved detection include, for example, spatially resolved illumination or a mixture of such methods, such as e.g. methods based on line-by-line scanning.
U.S. Publication No. 2004/0095360 A1 describes a user interface of an image recording device and an image processing method which can for example be used to evaluate biological samples such as pregnancy tests or drugs tests. In the process, a high-resolution camera sensor, designed as a color sensor, is used for actually capturing the image. Inter alia, it is proposed here that use is made of a line and a reference line within the test.
U.S. Pat. No. 7,344,081 B2 describes a method for automatically recognizing a test result of a sample zone on a test strip. In the process, an image of a barcode and an image of at least one test strip are recorded. A color response of the test strip to a sample application is determined. However, in order to resolve a barcode, it is inherently necessary to use a detector with a high resolution and hence with a large number of pixels.
U.S. Pat. No. 5,083,214 describes a device and a method for determining suitable points to take a sample. In the process, an array detector is scanned over a code present in the form of a microfilm, with a reduction in the data to be recorded being achieved as a result of the specific type of encoding. A challenge in this method consists of recognizing moving parts and, in the process, of in particular capturing information in digital form.
German Patent No. DE 196 31 086 A1 has disclosed an active pixel image sensor row, which uses protective rings, protective diffusions or a combination of these two techniques in order to prevent electrodes generated at the edge of an active region from being incident on an image sensor matrix. U.S. Publication No. 2007/0046803 A1 has disclosed a CMOS image sensor with a plurality of active pixel rows and one optically black pixel row. The optically black pixel row is activated in order to generate a respective optically black signal when each of at least two of the active pixel rows are activated. Both documents relate to specific aspects of the chip design of optically sensitive chips.
However, known approaches for evaluating images of test elements have the problem that an image has to be detected with a comparatively high resolution and has to be analyzed, for example using pattern recognition methods. In this case, a comparatively high resolution for example means a total of 1 million pixels, with, however, pixel arrays with a smaller number of pixels also being used in principle. Nevertheless, it is still necessary to transmit a large amount of data to peripheral electronics within a short space of time, e.g. 100 ms, and to use the latter to evaluate said data online; this significantly restricts the service life of the battery, particularly in the case of portable, e.g. handheld, instruments due to the high clock rates of the electronics required for this and due to the large number of computational operations. A partial solution is offered by pre-processing the image information in peripheral electronics. Such methods and devices are described in, for example, European Patent No. EP 1 351 189 A1, U.S. Publication No. 2005/0013494 A1 or in U.S. Publication No. 2003/0123087 A1. Alternatively, pre-processing in part already lends itself directly to a CMOS sensor, as described in e.g. U.S. Pat. No. 6,515,702 B1.
A method of a histogram evaluation was proposed as an alternative to a conventional on-chip or off-chip pattern recognition which subdivides the image into wetted, i.e. glucose-information carrying, and unwetted regions for further analysis. This method is described in European Patent No. EP 1 843 148 A1. Here, a frequency distribution is established for the detected light intensities, with the frequency distribution having at least a first maximum caused by unwetted portions and a second maximum caused by wetted portions. The concentration of the analyte is established from the frequency distribution. However, even though the histogram analysis, which for example is implemented directly on the CMOS image sensor, would significantly reduce the amount of data to be transmitted and evaluated, the proposed method, if preceding image preprocessing is to be avoided, realistically would still require significantly more than 10 000 pixels in order thus to enable a sufficiently precise glucose measurement.
In addition to the analysis and proposals outlined above, there moreover is the requirement from the point of view of metrology that the dimensions of typical measuring instruments are to be kept very small, which leads to significant consequences in terms of the flexibility of the optical unit layout. With increasing miniaturization, the lens used to image a test spot on a detector must have an increasingly higher refractive index, leading to increasing aberrations. By way of example, this results in the imaging being out of focus at the edges. In order not to compromise the image quality any further, the smallest possible pixels are desirable from an optics point of view. At the same time, the pixel dimensions on the detector are, as a result of the semiconductor processing technology of such sensors, generally restricted to values of at least 4 μm, of preferably more than 8 μm and of particularly preferably more than 20 μm. This means that semiconductor technology in general requires pixels that are as large as possible, whereas the optical unit layout requires ones that are as small as possible. This conversely leads to the necessity of magnifying imaging and, as a result, to a further reduction in the imaging quality in practical, cost-effective systems. Furthermore, the requirements on the positional tolerance of the test field increase in conventional systems with an increasing magnification scale.